تفاوت های جنسیتی در تعاملات آنلاین کودکان نوجوان:حالت های نمادین از خودارائه گری و خوداظهاری

Abstract Preadolescent children who did not know one another interacted in a multiuser domain (MUD), an online site designed to facilitate identity exploration and peer interaction. Each child participated in two separate sessions, one with a same-sex and one with an opposite-sex peer. Children created characters that reflected real-life properties of themselves, such as gender and interests in popular culture. Boys in same-sex pairs interacted with one another through action, rapid changes, and playful exchanges. Girls in same-sex pairs interacted primarily through written dialogue. In mixed pairs, boys wrote more and engaged in less playful exchanges, and girls wrote fewer and increased their actions. The results suggest that boys and girls have their own unique play styles with same-sex peers, but will moderate those patterns during late childhood to communicate with peers of the opposite sex.

. Introduction The self includes representations of the physical and psychological characteristics of who we are, as well as an ideal self of who we want to become (Harter, 1998). Physical characteristics of the self revolve around the body in which we live, a body that is constrained by attributes such as age, sex or gender, and physical appearance (Harre, 1983). Psychological characteristics of the self include the activities we do, the social skills we possess, and the roles that we play (Harter, 1998). How we present the self, characterized by its physical and psychological features, is the topic of a large literature on self-presentation Harre, 1983 and Harter, 1998. In most considerations of the self, it is necessary to consider the match between the actual, or underlying self, and the presented self, sometimes called the persona (Hall & Nordby, 1973). Biological sex is one aspect that is often important in people's self-constructions (Ruble & Martin, 1998). In the sex-typing literature, distinctions are drawn between biological sex and gender, the latter being socially constructed (Huston, 1983). Gendered behaviors are a key marker of self-expression for many children and adults alike, providing one anchor of identity. Our gender often influences the way that we interact with others, and has been called an interpersonal aspect of identity (Baumeister, 1997). Gender-related behavior becomes especially salient as children approach adolescence where they move from a world dominated by same-sex interactions to one that increasingly includes opposite-sex peers. Online multiuser domains (MUDs), where players can assume fantasy roles and engage in role playing activities, offer an easy context for exploring the self Turkle, 1995 and Turkle, 1997. In anonymous online interactions, we are free to present ourselves in many ways, less constrained by the expectations of the real world (Calvert, 2002). Investigating such online explorations may offer a window on how children represent themselves to others by examining their symbolic modes of self-presentation and self-expression. In this exploratory study, sex differences in the interactions of 11- and 12-year-old boys and girls were examined. We created an online MUD that was designed to foster social interaction and role-playing activities. In our data collection and analyses, we focused on three features that can serve as gender-related identifiers in online MUD interactions: (1) users' pretend names; (2) the kinds of pretend characters they create; and (3) the kinds of interpersonal roles and activities they undertake in relation to one another. We considered the interactions of unacquainted children in same- and opposite-sex pairs to maximize participants' flexibility in constructing a gendered character as well as in how easy it was for them to act in ways that were consistent or different from traditional gender roles. 1.1. Social interactions in MUDs MUDs are online forums in which participants interact with one another (Turkle, 1995). These spaces were originally only a text medium but now allow interactions in an online, three-dimensional world. Players can play games and engage in somewhat constrained role play activities ranging from slaying dragons to pretending to be Barbie, or they can create environments and explore the MUD in more flexible, and open-ended ways (Turkle, 1995). Thousands of people, many of whom are adolescents, spend time online in these role playing activities (Turkle, 1995). In MUDs, players create personas, or public masks—called avatars—in which they construct names, genders, and self-descriptions (Curtis, 1997). Some players use this medium to try out new roles to practice skills, such as learning to be more assertive; some alter core aspects of their identity such as their sex to learn what it is like to be a member of the opposite sex; some vary other dimensions to pretend to be people that are different from them (Turkle, 1995). These symbolic experiences can contribute to identity construction by allowing players to suspend reality and experience interactions as someone else, particularly when they interact with strangers. Players may assume their roles and practice activities in a safe space where accidents can be repaired simply by creating a new online character (Calvert, 2002). Each of the features we investigated—names, characters, and online interactions—can provide rich information about self-presentation. 1.1.1. What's in a name? People's names provide a major identifier of who they are and affect how they are treated. Researchers have shown that name popularity affects social interactions: children with frequent and popular names like Tom and Mary are better liked than children with rare, unpopular names like Herman and Hilda (Asher, Oden, & Gottman, 1977). Popular names minimize social discomfort, making it easier for children to move through the world. As the adolescent years approach, children become more aware of fitting in, of not wanting to stand out and be too different from their peers. One's name at this point may take on more valence. Hence, when given a choice, the kinds of names that children might select may be ones that are frequent in the peer group and that reflect popular culture. Nicknames also serve a social function in the worlds of children. For instance, nicknames can serve to separate “us” from “them,” as children without nicknames tend to be socially isolated from their peers (Harre, 1980). 1.1.2. Who shall I be? Creating a virtual body offers a window into participants' salient personal dimensions. Gender is one of the most salient in self-presentations and in the social construction of the self (Ruble & Martin, 1998). But would children choose to be their own gender online, or would they explore other gender identities if they had a choice? Some research indicates that adolescents as well as adults often engage in gender-bending when online, trying out the virtual identity of the other biological sex (Turkle, 1995). Another marker of identity in online avatar construction is the costume chosen, which signifies a particular role or stance that one assumes in relation to others. The chosen character or role influences how others treat a person online as well as how that person acts in relation to other characters (Turkle, 1995). An athlete, for instance, has a particular role to fill just as a wizard does. Roles may be selected online to reflect the popular culture, or may be selected to experience interactions in different roles. 1.1.3. How do I interact with others? Online interactions allow children to experiment with and develop interpersonal aspects of identity (Baumeister, 1997). These kinds of experiences allow them to experiment with the consequences of different kinds of interactions. The interaction preferences of boys and girls are quite different in late childhood, reflecting their experiences in a gender-segregated peer group (Maccoby, 1998). Boys tend to prefer more active interactions in visual media—such as action-adventure television programs and video games where content moves fast and changes rapidly, features that are considered to be perceptually salient Huston et al., 1990 and Roberts et al., 1999. By contrast, girls tend to prefer nonsalient features such as written words and dialogue Calvert & Kotler, 2003, Huston et al., 1990, Kafai, 1996 and Roberts et al., 1999. Boys also engage in more play than girls do, including role play (Subrahmanyam & Greenfield, 1998; Strommen, E., 2003, June 13, personal communication), and boys appear to learn best from playful activities where they are able to enact content. For example, young boys understand important television content better after exposure to structured role play rehearsals whereas young girls learn better from verbal summaries of that content (Friedrich & Stein, 1975). Such findings suggest that there may be sex differences in learning styles. Specifically, boys seem more likely to engage in enactive representations, in which they do things with their bodies, as well as iconic representations, in which they visualize content; by contrast, girls seem more likely to engage in verbal, symbolic styles of interaction and representation. During preadolescence, children show greater interest in interacting with a broad range of people, and new motivations to understand and communicate with others emerge. For example, romantic interests develop and boys and girls begin to navigate the intricacies of interactions with the opposite sex. How do boys and girls make a transition from same- to opposite-sex interactions? Online interactions, we believe, can shed light into how preadolescents learn to interact with opposite-sex peers. 1.2. The present study The present study provides descriptive information about preadolescent interactions with same- and opposite-sex peers in an online MUD. We were particularly interested in the following questions: (1) how would children present themselves to one another online, and would those presentations vary with same- and opposite-sex peers? (2) What would children do with one another online? Would the kinds of preferences children show with traditional media, such as boys' interest in perceptually salient features like action and girls' interest in features like dialogue, also appear in online interactions? and (3) Would play patterns be similar to those observed in real life, or would they be different online?

3. Results 3.1. Avatar construction 3.1.1. Avatar name The names children selected for their avatars were classified into one of six categories: real (42%), nicknames (25%), pop culture (8%), fantasy/mythical (9%), concepts and objects (7%), sports (6%), and nonsense (3%). As seen in Table 1, boys and girls picked different kinds of names when they picked a nonreal name, χ2(6) = 26.98, p < .001. Among those children who did not choose a real name or nickname, boys chose fantasy and mythical names, sports names, concept names, and nonsense names whereas most girls (and no boys) chose musical pop culture names. Similar numbers of boys and girls selected real names or nicknames. Table 1. Avatar names created by preadolescent boys and girls (%) Types of avatar names Real Nickname Pop culture Myth/fantasy Concept/object Sport Nonsense Boy (n = 82) 41 21 0 16 9 9 5 Girl (n = 76) 42 29 17 3 5 3 1 Total (N = 158) 42 25 8 9 7 6 3 Table options 3.1.2. Avatar gender Children overwhelmingly selected a gender for their character that was consistent with their own gender, χ2(1) = 149.19, p < .001. Only two boys (one 5th grader and one 6th grader) gender swapped by creating a girl avatar. Both occurred in the second session. One of those boys immediately identified himself as a boy when he entered the session, even though his character was a girl. 3.1.3. Avatar costume The most selected costume/role for avatars was athlete (39%), followed by punk kid (24%), normal kid (17%), wizard (21%), and firefighter (12%). As seen in Table 2, there were sex differences in these choices, χ2(4) = 26.26, p < .001. Girls were more likely to select athletes for their avatars, and boys were more likely to choose punk characters. The least selected choices were those of firefighter and wizard, roles that are less linked to children's own daily lives. Table 2. Sex differences in the avatar roles selected (%) Character/roles Normal Punk Firefighter Athlete Wizard Boys (n = 82) 20 34 11 21 15 Girls (n = 76) 13 12 4 59 12 Total (N = 158) 16 23 8 39 13 Table options 3.2. Overview: analyses of online interactions Scores were computed individually for each child for each dependent variable. Individual scores were then analyzed by the type of participant pair (boy, girl, or mixed). Analyses examined how children interacted with one another, focusing on what they did (movement, scene changes, emotional expression changes, role play, and game play), and said (number of words communicated, use of coded language). Time spent moving, the number of emotions displayed, the number of scenes visited, the number of words communicated, the number of coded word phrases used, time spent in role play activities, and time spent in game play activities were analyzed, in turn, by a 3 (gender pair: boy, girl, or mixed pair) × 2 (grade: 5th vs. 6th) ANCOVA using session order and the length of individual sessions as covariates.2 The unadjusted means are presented in Table 3. Adjusted means (M) and standard errors (SE) are reported for significant effects within the text for the analyses. Table 3. Uncorrected mean scores for MUD activities for children in same- and mixed-sex pairs Boys—same (n = 40) Mixed (n = 84) Girls—same (n = 34) Total (N = 158) Duration spent moving (s) 22.63 11.93 4.4 13.02 Number scenes visited 15.05 5.47 6.46 8.51 Duration in specific scenes (s): Beach 123.93 130.73 194.15 144.38 Park 81.18 108.28 151.97 111.19 Stage 52.85 113.25 79.06 87.32 City 154.30 73.39 54.26 92.13 Castle 52.05 45.55 55.74 49.94 Space 104.45 140.95 66.53 112.04 Number of emotions expressed 13.58 8.69 8.53 9.89 Duration expressing specific emotions: Happy 388.58 498.14 511.38 473.20 Sad 21.30 4.69 7.00 9.39 Angry 20.50 7.12 4.85 10.02 Silly 76.50 59.52 44.74 60.69 Surprised 35.75 16.39 22.74 19.53 Bored 28.15 18.87 11.00 22.66 Number of words written 55.68 73.88 93.41 73.47 Number of coded words/phrases used 1.88 2.65 3.68 2.68 Time in role play (s) 32.95 10.69 1.65 14.38 Time in game play (s) 32.05 9.26 2.41 13.56 Table options 3.2.1. Avatar movement The two-factor ANCOVA computed on the duration of time in seconds that children moved their avatars around the screen yielded a main effect of gender pair, F(2,150) = 11.41, p < .001. As expected, LSD post hoc follow-ups revealed that children in boy pairs (M = 22.71, SE = 2.62) moved significantly more than children in mixed sex pairs (M = 11.79, SE = 1.84), who, in turn, moved significantly more than children in girl pairs (M = 4.64, SE = 2.92). 3.2.2. Scenes visited Children spent most of their time together within the same scene (M = 91.42%, SD = 13.50). Children in mixed-sex pairs (M = 95.57%, SD = 7.67) were more likely to stay within the same scene than were children in boy pairs (M = 88.28%, SD = 11.04) or children in girl pairs (M = 87.31%, SD = 21.24), F(2,68) = 3.01, p < .06. The 2 factor ANCOVA computed on the number (frequency) of scenes visited yielded a main effect of gender pair, F(2,130) = 12.90, p < .001. As expected, LSD post hoc follow-ups disclosed that children in boy pairs (M = 15.54, SE = 1.58) changed scenes more often than children in mixed sex pairs (M = 5.52, SE = 1.40) or children in girl pairs (M = 6.40, SE = 1.71). Descriptive statistics showed that children spent the most time overall at the beach (M = 23.90%, SE = 2.65), followed by the park (M = 19.38%, SE = 2.59), space (M = 18.07%, SE = 2.20), the city (M = 15.69%, SE = 2.14), the stage, which was the default option (M = 14.35%, SE = 2.23), and the castle (M = 8.61%, SE = 1.43), respectively. 3.2.3. Emotions expressed The 2 factor ANCOVA computed on the frequency of emotions expressed yielded a main effect for gender pair, F(2,150) = 4.42, p = .01, which was qualified by a gender pair by grade interaction, F(2,150) = 4.99, p < .01. LSD post hoc follow-ups revealed that children in boy pairs (M = 13.69, SE = 1.51) changed emotions more frequently than did children in mixed-sex pairs (M = 8.85, SE = 1.06) or children in girl pairs (M = 7.89, SE = 1.69). As seen in Fig. 2, the Gender Pair × Grade interaction revealed that gender pair differences occurred for 6th, but not 5th graders; specifically, 6th grade children in boy pairs changed scenes much more often than children in mixed-sex pairs or children in girl pairs. Number of emotions selected as a function of pair type and grade. Fig. 2. Number of emotions selected as a function of pair type and grade. Figure options Descriptive statistics revealed that children spent the most percentage of time using happy expressions (M = 79.58%, SE = 2.03), which was the default choice, followed by silly (M = 10.23%, SE = 1.36), surprised (M = 4.00%, SE = 0.85), bored (M = 3.74%, SE = 0.99), angry (M = 1.70%, SE = 0.41), and sad expressions (M = 1.63%, SE = 0.33), respectively. 3.2.4. Frequency of online role play and game play activities Role play occurred in 27% of the sessions, as did game play. There were significant differences in the number of sessions that children in boy pairs, girl pairs, and mixed pairs participated in role play, χ2(2) = 15.36, p < .001 and in game play, χ2(2) = 14.24, p = .001. Play occurred more for boy pair than mixed pairs or girl pairs (45% of children in boy pairs showed role play; 17% of children in mixed pairs; 12% of children in girl pairs), as did game play (45% of children in boy pairs; 26% in mixed pairs, 6% in girl pairs). 3.2.5. Amount of time involved in online role play activities The two-factor ANCOVA computed on the duration of time in seconds that each child was involved in role play yielded a main effect of gender pair, F(2,150) = 4.59, p = .01; grade, F(1,150) = 5.72, p < .02; and a Gender Pair × Grade interaction, F(2,150) = 3.32, p < .05. As expected, children in boy pairs (M = 33.01, SE = 7.46) spent significantly more time engaged in role play than children in mixed-sex pairs (M = 10.51, SE = 5.24) or in girl pairs (M = 1.97, SE = 8.31), and 6th graders engaged in more role play than 5th graders (M = 25.46, SE = 5.84 vs. M = 4.87, SE = 5.96). As seen in Fig. 3, the Gender Pair × Grade interaction revealed that the gender pair differences in role play were more pronounced for 6th than for 5th graders, particularly for children in boy pairs. Amount of role play as a function of pair type and grade. Fig. 3. Amount of role play as a function of pair type and grade. Figure options 3.2.6. Amount of time involved in online game play activities The 2 factor ANCOVA computed on the duration of time in seconds that each child was involved in game play yielded a main effect of gender pair, F(2,150) = 15.38, p < .001. The session order covariate was also significant, F(1,150) = 4.31, p < .05. As expected, children in boy pairs (M = 33.47, SE = 4.18) spent significantly more time engaged in game play than children in same-sex pairs (M = 7.52, SE = 2.94) or children in girl pairs (M = 4.73, SE = 4.66). 3.2.7. Words communicated The two-factor ANCOVA computed on the total number of words communicated yielded a main effect of gender pair, F(2,150) = 14.07, p < .001. The covariate of session length was also significant, F(1,150) = 4.08, p < .05. As expected, LSD post hoc follow-ups revealed that children in girl pairs (M = 95.04, SE = 5.32) wrote more than children in same-sex pairs (M = 72.35, SE = 3.36), who in turn, wrote more than children in boy pairs (M = 57.47, SE = 5.32). 3.2.8. Coded language The two-factor ANCOVA computed on coded language scores yielded a main effect of grade, F(1,150) = 10.93, p = .001. Sixth graders used more coded language than fifth graders (M = 4.08, SE = 0.51 vs. M = 1.60, SE = 0.52, respectively). The kinds of coded language that children used included lol (laughing out loud), 2 (to or too), 18er (later), nvm (never mind), g2g (got to go), ur (you are), and lylas (love you like a sister). We also observed a child explaining the definition of lol to another child, suggesting that they explicitly teach each other the meaning of coded language.